Precipitation Regulates the Response of Net Ecosystem CO2 Exchange to Environmental Variation on United States Rangelands

Abstract

Rangelands occupy 50% of Earth’s land surface and thus are important in the terrestrial carbon (C) cycle. For rangelands and other terrestrial ecosystems, the balance between photosynthetic uptake of carbon dioxide (CO2) and CO2 loss to respiration varies among years in response to interannual variation in the environment. Variability in CO2 exchange results from interannual differences in 1) environmental variables at a given point in the annual cycle (direct effects of the environment) and in 2) the response of fluxes to a given change in the environment because of interannual changes in biological factors that regulate photosynthesis and respiration (functional change). Functional change is calculated as the contribution of among-year differences in slopes of flux-environment relationships to the total variance in fluxes explained by the environment. Functional change complicates environmental-based predictions of CO2 exchange, yet its causes and contribution to flux variability remain poorly defined. We determine contributions of functional change and direct effects of the environment to interannual variation in net ecosystem exchange of CO2 (NEE) of eight rangeland ecosystems in the western United States (58 site-years of data). We predicted that 1) functional change is correlated with interannual change in precipitation on each rangeland and 2) the contribution of functional change to variance in NEE increases among rangelands as mean precipitation increases. Functional change explained 10-40% of the variance in NEE and accounted for more than twice the variance in fluxes of direct effects of environmental variability for six of the eight ecosystems. Functional change was associated with interannual variation in precipitation on most rangelands but, contrary to prediction, contributed proportionally more to variance in NEE on arid than more mesic ecosystems. Results indicate that we must account for the influence of precipitation on flux-environment relationships if we are to distinguish environmental from management effects on rangeland C balance.